Persistent Systems Ltd. hopes to give healthcare a new lease on life with its new life sciences solution, Pi-OmniKG.

The modular Knowledge Graph, or KG framework, could help life sciences professionals contend with the vast amount of data involved in that industry and make new discoveries.

Persistent’s Santosh Dixit and Google Cloud’s Shweta Maniar, talk with theCUBE about how AI and knowledge graphs could impact the life sciences industry.

“During the COVID pandemic, there was a big explosion of data … where the need for converting data into actionable insights became more powerful,” said Santosh Dixit (pictured, right), chief domain expert at Persistent. “The biological truth lies in bringing all this data together, and … this is something we have taken the help of a concept in [artificial intelligence] called knowledge graphs. With the help of a knowledge graph construct with the Google tech stack, we have created this unique solution called Pi-OmniKG.”

Dixit and Shweta Maniar (left), global director of healthcare and life sciences at Google Cloud, spoke with theCUBE’s Rebecca Knight at theCUBE’s Coverage of Google Cloud at HIMSS25, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed how AI and knowledge graphs could impact the life sciences industry. (* Disclosure below.)

Making sense of life sciences data with Pi-OmniKG

Life sciences data poses a particular problem on the software level because it is enormous, interconnected and multimodal. Pi-OmniKG offers a solution by uniting disparate types of data, structured and unstructured, into a knowledge graph, though the data needs to be curated in order to get good results, according to Dixit.

“Typically in life sciences, you are looking at senior-level scientists who want to integrate a lot of data sources coming together to derive what is called actionable intelligence,” he said. “And that actionable intelligence resides in good quality data, which is upstream. Knowledge graphs were always popular, but in the last 18 months, the ability to have a [large language model] talk to a knowledge graph has probably given it a second wind.”

Once the knowledge graph is established, users can then interact with it using an LLM. One feature that Google and Persistent are working on is getting visual output as well as text from a knowledge graph, giving researchers and scientists an even more accessible picture of their data, according to Maniar.

“Having trust in these systems, even if you’re not a technical expert, is going to be key to how we’re going to be able to move forward,” he said. “Those who are going to be deploying, for example, Pi-OmniKG, may not be technical experts … but they’re coming with these science problems and a lot of data, and they’re looking at how do you derive value from this information by asking the right questions of the system.”

There are numerous use cases for this technology, including drug discovery, since a knowledge graph can visualize the relationships between genomic data in a new way. Precision medicine, a type of care tailored to the individual patient, is another key application for Pi-OmniKG, according to Dixit.

“You want to have a bullseye approach to treating a patient,” he said. “We … have a use case of Pi-OmniKG where we look at a genomic profile of a lung cancer patient, and we try to ask [the] question, ‘How can I identify the right choice of the drug for this kind of a genomic profile?’ All this is driven by bringing in a lot of data sets associated with precision oncology.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s Coverage of Google Cloud at HIMSS25:

(* Disclosure: TheCUBE is a paid media partner for theCUBE’s Coverage of Google Cloud at HIMSS25. Neither Google LLC, the sponsor of theCUBE’s event coverage, nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE